Table Of Contents
Release Notes for CiscoWorks Network Compliance Manager, 1.2.1
NCM and LMS Co-residency Requirements
NCM High Availability System Requirements
Installing the Enhanced PERL API
Change Detection and Configuration Management
Policy Manager and Compliance Violations
Batch Changes and Task Management
Known Limitations and Problems
Administrative Settings - User Authentication Page Crypto Key Exception
Banner Handling Strings Require Device-specific Passwords
BayRS Device Can Lose Ability to Provide Snapshot
BayStack 450 Could Stop Responding to Telnet, SNMP, or ICMP
Cisco Banner Messages Special Characters
Command Line Interface: connect Command
Command Line Interface: Set Telnet or SSH Client Width to 500
Console Server: SSH Access is not Supported
Deploy to Startup Config and Reboot not Supported Using SNMP
Diagnostics: When to Run ICMP Tests
Distributed System Performance
Duplicate IP Addresses with Multiple Sites
Extreme Devices: Configuration Comments Can Cause Misconfiguration
Inventory: Data from Device Overwrites Manually Entered Values
Juniper Devices with SCP Enabled do not Capture Running Configurations
RADIUS External Authentication
Reports: Checkpointing Can Cause Reports to be Inflated
Scripts: Cannot Save Command Scripts with Quote Marks in the Name
Scripts: Cannot Save Template or Command Scripts with a Period in the Name
Scripts: Command Scripts and Templates for Cisco Aironet VxWorks Devices
Scripts: Output Results in HTML Format
SecurID Software Token Software, Version 3.0.5
Sending Reports to External Email Addresses
Software Center: Cisco IOS 2500
Software Center: Cisco IOS Devices
Software Center: Deploying Software
Software Center: Downgrading Nortel OS and Rebooting Could Leave Device Inaccessible
Software Center: Image Set Name Requirements
Software Center: Reboot Option
SQL Server 2005 Password Requirement
Tasks: Running External Application Tasks Presents a Possible Security Risk
Tasks: Task Scheduled for the 31st Might Run on the 1st
Unresponsive Script Warning Message in Mozilla Firefox 1.5 (or Higher)
Obtaining Documentation, Obtaining Support, and Security Guidelines
Release Notes for CiscoWorks Network Compliance Manager, 1.2.1
August, 2007
These release notes include important information regarding CiscoWorks Network Compliance Manager (NCM), Release 1.2.1. NCM provides network configuration and change features, policy-based workflows, best of class compliance reporting capabilities, and APIs. NCM includes integration with CiscoWorks—initially launchable from the CiscoWorks home page and interoperability with other CiscoWorks applications such as the LMS bundle through the CommonServices Device Credential Repository (DCR).
Note
All documentation, including this document and any or all of the parts of the NCM documentation set, might be upgraded over time. Therefore, we recommend you access the NCM documentation set using the Cisco.com URL: http://www.cisco.com/en/US/products/ps6923/tsd_products_support_series_home.html.
In addition, the Docs tab visible from within Network Compliance Manager might not include links to the latest documents.
Contents
This release note contains the following sections:
•
Known Limitations and Problems
•
Obtaining Documentation, Obtaining Support, and Security Guidelines
Features in This Release
NCM Release 1.2.1
The NCM 1.2.1 release includes the following enhancements:
•
Keyboard Interactive Authentication (KBI) is now a supported method for SSH connections to devices. This method is commonly used by F5 devices. Now, you can manage F5 devices via NCM that require KBI authentication.
•
To help manage the volume of Security Alerts, the Software Compliance page now displays Last Modified and CVE columns that you can sort. To view this enhancement:
–
Navigate Policies > Software Compliance.
–
Click View > Security Alert Service Alerts
–
Click the Last Modified column. The most recent Security Alerts downloaded from the Cisco Network are displayed at the top of the list.
•
NCM now provides a PING and Traceroute command to test NCM server-to-device connectivity. This command is provided in the Telnet/SSH Proxy and CLI. To learn more about how to use these commands, go to the NCM command line and run help os ping and help os traceroute.
•
NCM now provides a List Imageset CLI/API command to show the list of image sets present in NCM. To learn more about how to use this command, go to the NCM Command Line and run help list imageset.
•
When all tasks are configured to use AAA credentials, snapshot tasks triggered by change events can now be configured to use the AAA credentials of a known user in the case where the task would run without a known user. To learn more about how to use these commands, go to the NCM Command Line and run help os ping and help os traceroute.
•
The Perl and SOAP APIs now support the Import command. The Import command enables you to import devices into NCM. To learn more about the Import command in the APIs, go to this location <NCM Install Directory>/docs/ and open the SOAP_API_Guide.html document. Instructions on the new Import API are provided.
•
Users can now configure NCM to automatically use a specific user to retrieve configuration if no AAA credentials are available. If NCM is configured to only access devices via AAA credentials and NCM detects a change by a user who does not have AAA credentials defined, it will failover to use the Admin credentials to access the device and collect the snapshot. In NCM 1.2.1, users can now specify a specific user account they want NCM to use in this failover situation.
To configure this new setting, you must have NCM configured to use AAA credentials for snapshot tasks. To do this, navigate Admin Settings >Device Access >Task Credentials and uncheck the Allow Standard Device Credentials option for Take Snapshot. Then, check the corresponding option for Allow User AAA Credentials for Take Snapshot.
To enable a specific user account to be used in cases where NCM does not have AAA credentials to collect a snapshot, do the following:
a.
Navigate Admin Settings >Device Access >Task Credentials.
b.
At the bottom of this section, in the Fallback Admin User field, enter the Username you want NCM to use to take the snapshot.
Note
Note: There is no validation that the username you entered is correct. If you add an invalid Fallback admin user, the snapshot will fail and provide a descriptive error message.
•
The following commands have been added to the new Perl and SOAP APIs. To learn more about these command in the APIs, login to NCM and click the Docs link. Click the SOAP API Reference Guide for Network Compliance Manager link.
–
add image
–
add metadata
–
add metadata field
–
add partition
–
add user to group
–
delete image
–
del metadata
–
del metadata field
–
del partition
–
del user from group
–
deploy image
–
import
–
list all drivers
–
list config id
–
list device family
–
list device group
–
list device id
–
list device model
–
list device software
–
List device type
–
list device vendor
–
list image
–
list imageoption
–
list imageset
–
list metadata
–
list metadata field
–
list partition
–
list script id
–
list script mode
–
list site
–
list topology
–
list topology graph
–
list topology ip
–
list topology mac
–
list view
–
list vlan
–
list vlan ports
–
login
–
logout
–
mod metadata
–
mod metadata field
–
mod partition
–
mod topology graph
–
run gc
–
show device family
–
show device latest diff
–
show metadata
–
show metadata field
–
show server option
–
show topology
–
show version
–
stop task
–
stop task all
–
test view
–
undeploy image
–
update dynamic group
•
The Auto-remediation page in NCM now provides warning text. Because auto-remediation starts immediately when a violation occurs, Opsware recommends you enable Workflow and Approvals when the Auto-remediation task runs.
NCM Release 1.2
The 1.2 release includes the following features:
•
Automated cisco.com product license registration, which enables:
–
Flexible licensing
–
Incremental node license registration
–
Incremental feature license registration for features such as high availability, satellite, and connectors
•
Optional subscription to NCM Alert Center content packs, which can be downloaded from the cisco.com NCM Alert Center Web page into the NCM application. These content packs can keep you up to date with:
–
security compliance policies
–
product extensions
•
Subscriber access to and ability to download content packs into NCM.
•
New connector with third-party network management products such as Remedy AR.
System Requirements
This section includes the following:
Linux Server Requirements
The following tables provide the recommended requirements when installing NCM on a Linux platform. Keep in mind that the application server and the database server can be configured together or separately depending on the size of the network.
Note
You must stop other network management applications, Web servers, databases, and Syslog/TFTP servers running on the same system before installing NCM. Applications include anti-virus (during Setup only) and WWW Publishing Server applications.
Note
When installing NCM on a Linux platform, Nmap 3.81 is required for Nmap scanning when running the Detect Network Devices task.
Summary Reports
Summary reports are generated in the Microsoft Excel XLS format. Excel does not run on Linux. You can either run the Summary reports from a Windows client computer connected to your NCM server or you can use one of the following products that run on Linux and can open Excel files:
•
Open Office (www.openoffice.org)
•
GNUmeric (www.gnumeric.org)
•
Star Office (wwws.sun.com/software/star/staroffice)
Solaris Server Requirements
The following tables provide the recommended requirements when installing NCM on a Solaris platform. Keep in mind that the application server and the database server can be configured together or separately depending on the size of the network.
Note
You must stop other network management applications, Web servers, databases, and Syslog/TFTP servers running on the same system before installing NCM. Applications include anti-virus (during Setup only) and WWW Publishing Server applications.
Note
When installing NCM on a Solaris platform, Nmap 3.81 is required for Nmap scanning when running the Detect Network Devices task.
Summary Reports
Summary reports are generated in the Microsoft Excel XLS format. Excel does not run on Solaris. You can either run the Summary reports from a Windows client computer connected to your NCM server or you can use one of the following products that run on Linux and can open Excel files:
•
Open Office (www.openoffice.org)
•
GNUmeric (www.gnumeric.org)
•
Star Office (wwws.sun.com/software/star/staroffice)
Windows Server Requirements
The following tables provide the recommended requirements when installing NCM on a Windows platform. Keep in mind that the application server and the database server can be configured together or separately depending on the size of the network.
Note
You must stop other network management applications, Web servers, databases, and Syslog/TFTP servers running on the same system before installing NCM. Applications include anti-virus (during Setup only) and WWW Publishing Server applications.
Note
When installing NCM on a Windows platform, Nmap 3.81 and WinPcap (Windows Packet Capture Library) version 3.1 are required for Nmap scanning when running the Detect Network Devices task.
NCM and LMS Co-residency Requirements
The following are the recommended requirements when you are enabling co-residency of NCM and CiscoWorks LAN Management Solution (LMS):
•
Operating System on the Application Server: Microsoft Windows 2003
•
Server Hardware: At least a Xeon (or a Dual Core) Processor with 8 GB of RAM.
For detailed information on NCM and LMS co-residency, refer to the Configuration Guide for Network Compliance Manager and LMS Co-residency.
Installation Considerations
Note
We sometimes release patches after the original release of a product. Therefore, you should review the the following site on Cisco.com for any updates.
http://www.cisco.com/cgi-bin/tablebuild.pl/cwncm-crypto
NCM Gateway Requirements
The Network Compliance Manager Gateway enables a NCM Core to manage servers that are behind one or more NAT devices or firewalls. The NCM Gateway is supported on the following platforms:
•
RedHat Linux 3.0 AS
•
RedHat Linux 4.0 AS
•
SuSE Linux 9.0 ES
•
SunOS 5.9
•
SunOS 5.10
NCM High Availability System Requirements
The NCM High Availability Distributed System is a multi-master system where the data from each NCM Core is available to all other NCM Cores. This collection of NCM Cores is called an NCM mesh. This configuration helps provides a comprehensive view of your data and allows for redundant data and failover in the event of a problem with the NCM Core. Each NCM Core consists of an NCM Management Engine, its associated services (Syslog and TFTP), and a single database.
Note
If you intend to install the NCM High Availability Distributed System, keep in mind that it only supports the Oracle 9.2 database server. If you are running Oracle 10.2 on the Core database server, you cannot upgrade to the NCM High Availability Distributed System.
Upgrading the OS
Before performing server maintenance or upgrades, take steps to ensure that in the case of application or OS corruption, you will be able to restore critical NCM data.
Step 1
Back up your database to a safe location before doing server maintenance.
Step 2
Save the contents of the following folders to a safe location:
<NCM directory>\client\scripts\
<NCM directory>\addins\
<NCM directory>\jre\
<NCM directory>\server\lib\drivers
<NCM directory>\docs
If upgrading the server causes NCM to malfunction, contact Technical Support.
Backing Up User Files
During an upgrade, the NCM Setup program automatically backs up user files, such as the Summary reports and NCM scripts, to the following directories:
For Windows:
\winnt\temp\rendition\addins
\winnt\temp\rendition\images
\winnt\temp\rendition\log
\winnt\temp\rendition\scripts
For Solaris:
/var/rendition/addins
/var/rendition/images
/var/rendition/log
/var/rendition/scripts
Note
If you have changed directories, you can find your current Windows directory by opening a command prompt window and entering set windir. The Windows home directory is displayed. The installer restores all user files automatically, except log files. If you want to keep appending to saved log files, copy them to \rendition\server\log.
Installing the Enhanced PERL API
The following modules are provided on the NCM Distribution CD:
•
Cisco::NCM::Util
•
Cisco::NCM::Client
•
Cisco::NCM::Connect
Installation Requirements
PERL version 5.8 or later is required.
If you are using the Auto Installer, skip to the "Auto Installer Method:" section.
Note
NCM 1.2 does not currently support the system call to run the install.pl script. At this time, you will have to manually install the enhanced PERL API.
If you are manually installing the PERL API, confirm that certain versions of PERL and/or PERL modules (that are not part of some core PERL distributions) are installed before you begin. Refer to the META.yml file within each package/tarball for its requirements.
If your PERL distribution does not contain all of the required PERL modules, they are available at http://www.cpan.org and/or via PPM. If you are using ActivePerl, try PPM first.
PPM (ppm.exe) is part of the ActivePerl distribution. If you are using ActivePerl, it is recommended that you use the PPM method. You can also run PPM without arguments and then issue the install command. You may need to do this for some PERL modules that have multiple versions to choose from, followed by install # (where # is the item in the list returned by the install command). Keep in mind that PPM prefers to use the - as a namespace separator in place of the PERL :: separator.
To install any of the required modules, use one of the following commands:
•
ppm install SOAP-Lite
•
cpan install SOAP::Lite
Note
NMAKE.EXE is installed when installing NCM on a Windows platform. It is located in the /client directory. CPAN is simply a wrapper for the PERL -MCPAN -e shell command. The CPAN command cpan.exe is part of the core PERL install on all PERL versions since 5.8.0, including ActivePerl.
Installation Steps
There are two methods for installing the PERL API modules. The first and easiest method is to use the Auto Installer. You can only use the Auto Installer, however, if you have installed the PERL API distribution via the NCM installer. Otherwise, you must use the manual installation method.
Auto Installer Method:
The Auto Installer installs all of the Cisco::NCM modules as well as their dependencies.
Step 1
Open a shell.
–
If you are on a Windows platform, open a command shell.
–
If you are on a Linux or Solaris platform, you can either open a command shell or SSH into the NCM server.
Note
You will need to have privileges to both create and modify files for NCM as well as PERL. As a result, you might need Administrator privileges on a Windows Platform and root privileges on Linux or Solaris platforms.
Step 2
Change to the directory where NCM is installed. This directory will have been set when you installed NCM.
Step 3
To run the install script, enter:
perl client/perl_api/har/install.pl
Note
If PERL is not in your path or you have multiple PERL versions installed, use the full path to the PERL executable that you will be using. This should also match the value for the PERL interpreter set in the NCM server configuration.
All of the Cisco::NCM modules are installed, as well as their dependencies. However, only pure PERL dependencies are provided. For example, SOAP::Lite is provided, which includes a minimalist lightweight XML parser. For the best performance, we recommend that you have the XML::Parser module installed.
If you are using ActivePerl (with a PERL version of 5.8 or better), the XML::Parser module is included with the distribution. Otherwise, you will need to use PPM, CPAN, or manually download and install the module.
Manual Install Method:
Keep in mind that the installation could fail if your PERL installation does not meet certain requirements. Refer to the "Installation Requirements" section. In addition, the Cisco::NCM PERL modules are distributed as compressed tarballs, similar modules on CPAN. They are located in the following directory: <NCM_ROOT>/client/perl_api/Cisco/.
To untar and uncompress all of the modules at one time, you can use the ptar command. ptar is distributed as part of the popular PERL module Archive::Tar, which is included in the standard ActivePerl distributions. To view the contents of the directory and to extract the contents into your current directory, enter: ptar -xzvf PATH/TO/whatever.tar.gz.
For each of the following modules, uncompress and untar the module(s) and change to the directory that was created:
•
Cisco::NCM::Util
•
Cisco::NCM::Client
•
Cisco::NCM::Connect
To install the PERL API on a Windows platform with ActivePerl, or any platform running a version of PERL that has the Module::Build module installed, enter:
•
perl Build.PL
•
perl Build build
•
perl Build test
•
perl Build install
You may also use the traditional CPAN method. Enter:
•
perl Makefile.PL
•
make
•
make test
•
make install
Note
If you are using the CPAN method on a Windows platform, you will need to enter nmake rather than make.
PERL Documentation
After installing the PERL API, you can view the following PERL POD pages:
•
perldoc Cisco::NCM::Client
•
perldoc Cisco::NCM::Connect
•
perldoc Cisco::NCM::Client::4_5_x
•
perldoc Cisco::NCM::Client::6_0_x
Your PERL distribution can also build HTML files for the documentation.
Examples
There are PERL API examples in the demo directory. These examples illustrate how to use the PERL API. Keep in mind that it is possible to run the examples without installing the PERL modules by remaining in the demo directory and supplying the relative (or full) path to each example, as in:
•
unix_box$ perl demo/list_users.pl
•
C:\Windows\Box> perl demo\list_users.pl
Resolved Problems
NCM 1.2.1 includes the following fixes:
Discovering Network Devices
Detect Network Devices task no longer adds new devices with network and broadcast addresses
In previous releases, when the Detect Network Devices task was run with a CIDR range, NCM would add devices for the network and broadcast addresses. NCM has been enhanced to remove the first and last address from a scan when the address scan is specified as a CIDR range. This eliminates the broadcast and network range from being scanned, and erroneously adding new devices.
Change Detection and Configuration Management
Edit & Deploy Configuration feature now correctly tracks the most recent configuration on a device
When using the Edit & Deploy Configuration feature, previously the configuration scheduled for deployment would erroneously show as the most recent configuration. This has been fixed.
Change Notification email now attributes the correct username
The Change Notification email from NCM no longer attributes the wrong username with configuration changes.
BayRS drivers no longer fail to parse binary configurations if circuit name starts with 0
The BayRS driver no longer fails to parse the binary configuration if a circuit name contains a number that starts with 0.
Parsing of syslog messages from Cisco devices improved for real-time change detection
In previous releases, if the msec in the timestamp of the syslog message was less than 255, NCM change detection failed. This has been fixed.
NCM now correctly displays more than one configuration in the configuration audit trail
In NCM 1.2.1, it was possible for NCM to only display one configuration for the device configuration audit trail, even if the device had many historical configurations. This has been fixed.
Searching and Reporting
Compliance Center reports now correctly list port availability
NCM Compliance Center reports now correctly report Port availability. Previously, the Free Ports column was not always accurate.
Advanced Search now contains the current number of context lines in mailed and CSV reports
Advanced Search reports now display the correct number of context lines in mailed and CSV reports based on search criteria.
Summary reports displayed last run date now matches filestamp date
The Summary reports, Summary tab, and Last Run: date now report the actual date when a report was run.
Links to Summary reports from Compliance Center and User & System reports now work correctly
In previous releases, the Compliance Center reports and the User & System reports had invalid links to the Summary reports.
Context lines are not maintained across multi-page search results
In previous releases, when searching for configurations that contain specific text, the option to specify the number of context lines did not work across pages in a multi-page search result display. As a result, when the Next Page button is selected to move to the next page of a multi-page search results display, the previously entered number of context lines reversed back to the default value of 3. This has been fixed.
The Application Switch device type is now a searchable option
The Application Switch device type is now listed as an option when searching for a device by device type.
Users are now able to run the Diagramming report from My Favorite reports
In previous releases, you could not run Diagramming reports from My Favorite reports. This has been fixed.
Device search for `Ports in Use' no longer fails with SQL error
In a previous release, a Search for Devices where the criteria included Ports in Use failed with a SQL error. This has been fixed.
Policy Manager and Compliance Violations
Configuration Policy Activity page now shows the most serious importance rating for a device with multiple compliance violations
The Configuration Policy activity page now shows the most serious importance rating for devices with multiple compliance violations. For example, if a device has a Medium and a Critical violation, the rolled up report shows the device as Critical.
Dynamic device group criteria based on device compliant state now works
For dynamic device groups where the filter is defined as "non-compliance with specific policy rules," the filter now works as expected.
Batch Changes and Task Management
Run Script command in the CLI/API now runs correctly with double-quotes
The Run Script command in the CLI/API now runs correctly when double-quotes are used. For example, previously Run Script would fail if double-quotes with a single character were used.
Software Update task page refresh no longer results in a page error
When updating a Software Update task, upon refresh, an error would occur: failed to save task: $1$. This has been fixed.
Checking the Show Task Detail checkbox no longer breaks multi-page task browsing
In previous releases, when viewing a list of Child tasks (subtasks) of a Group task, if you select the Show task detail checkbox and then move to the next page of the task, the page is not displayed. This has been fixed.
Saving a new template with more than 4,000 characters no longer returns an error
In previous releases, saving a new template with more then 4,000 characters returned the error: ORA-01704: string literal too long. This only impacts Oracle installations. This has been fixed. [
Event Rules
Event Rule emailing contents of a diagnostic task no longer fails to include actual diagnostic information
In previous releases, the event variables for diagnostics, $CurrentDiag$, $PreviousDiag$, and $Diff, were only valid for the Device Diagnostic Changed event. These event variables are now available in Event Rules from any Diagnostic event (Device Diagnostic Changed, Device Diagnostic Completed Successfully, and Device Diagnostic Failed).
Permissions
View Partition device view abilities enhanced
You can no longer set View Partition permissions to users for View Partition Permissions that they have not been granted or set View Partition Permissions on Partitions that they cannot view.
Users can no longer run scripts from the CLI/API without explicit permission
Users can no longer run scripts from the CLI/API without being granted Script Permission.
Granting a user Command Permission to "Run External Application" now enables this capability as an Event Rule for that user
Granting a user Command Permission to Run External Application now enables this capability as an Event Rule function for that user.
Performance
Event Pruner on Oracle can now delete more than 10K records
The Event pruner no longer stalls when deleting more than 10,000 records.
Single device tasks scheduled via the Device Selector are no longer erroneously scheduled as a group task
When using the Device Selector to schedule a task, if you schedule a task on a single device, NCM now schedules a standard single device task.
Batch editing on device groups
Performance has been improved by enabling batch editing on selected device groups (as opposed to using the select all option).
Search for devices no longer fails if search criteria equals more than 2,048 characters
You can now perform searches in NCM where the search criteria provided is greater than 2,048 characters. For example, Search for Devices where Device Group is <select long list of different device groups>.
Device Snapshot event filter no longer causes snapshot failures
In previous releases, enabling the Device Snapshot event filter in the Admin Settings caused snapshots to fail. This has been fixed. You can now disable this event to improve performance without any adverse effects.
Displaying large device list
When NCM displays a large device list, it now avoids a large CPU performance hit.
Syslog_wrapper.log file
The start_syslog_wrapper.log file now includes a pruning method to restrict file growth.
Security
Console Server password is now masked on output
The console server password is considered to be sensitive data and is therefore masked on output as {SECURE STRING}.
API/CLI
Perl API list-modules command now returns the correct type
The list_modules command in the NCM Perl API now returns the correct type and data.
Perl API list-ports command now returns the correct type
The list_ports command in the NCM Perl API now returns the correct type and data.
list_site command added to Perl API
The list_site command now exists in the NCM Perl API.
Table 10 lists the problems that were resolved in CiscoWorks Network Compliance Manager, Release 1.1. Table 11 lists the problems that were resolved in CiscoWorks Network Compliance Manager, Release 1.2.
Known Limitations and Problems
This section contains information about the limitations and problems known to exist in the NCM 1.2 product.
CSCse09644—The cwncm_import script does not parse the hostname as present in the CSV file.
Description: When exporting some devices from DCR into the CSV file using dclr_export.sh or from Device Management UI of LMS, the cwncm_import script does not parse the hostname (present) in the CSV file; instead, it substitutes the IP Address as the hostname for all these imported devices.
Workaround: You can manually change the Hostname by looking up the corresponding name in the CSV file. This issue will be fixed in a future release.
CSCse09092—Clicking Perform Test does not launch a new window.
Description: From Policies > Policy List > Add New rule > Test > Added Devices from Device Selector, select Perform Test. A browser window does not launch to display the Test Results screen.
Workaround: The problem only happens when you select a UNIX driver for the UNIX end hosts. Selecting any other device type will resolve the issue.
CSCse11820— Installation hangs if you provide incorrect Database credentials.
Description: Oracle is installed successfully and you proceed with NCM installation. If you provide any incorrect database credentials (port number, DB name, or password) while configuring the NCM Database, then NCM hangs while trying to connect to the database.
Workaround: Stop the installation using the Windows task manager. Restart the installation and enter the correct credentials.
CSCse14518—When you attempt to delete a large number of devices from the NCM database, it fails.
Description: Devices are imported into NCM, which uses Oracle database. Go to the Device > Inventory window, select a large number of devices that you wish to delete, and select Delete. NCM prompts you confirm the deletion. Click yes. After sometime, only 200 devices are deleted.
Workaround: Delete a smaller number of devices at each attempt (< 500 devices).
CSCse16371—An error message incomplete command displays when you try to get hardware information for a CRS-8/S device.
Description: From Inventory, select a CRS-8/S device. Go to View > Diagnostics > Hardware Information. No hardware info gets displayed, instead an error with the message incomplete command is displayed.
Workaround: There is no known workaround for this issue. Please avoid using this feature for this device type.
CSCse16848—Duplicate entries are seen in the software updates report.
Description: When adding more than one image set from Devices > Device tools > Software Images, the weekly report incorrectly reports two successful updates when this is not the case.
Workaround: There is no known workaround for this issue.
CSCsg79893—License monitor results column updated only after browser manual refresh.
Workaround: This is how all the System Monitors work. By default the monitor data is updated every 6 hours but this is configurable.
CSCsh28136—Installer fails to copy licenses from a directory whose name has spaces.
Workaround: Make sure that the directory and directory path where the license files are being copied do not have spaces in their names. If you must use directory and directory path names containing spaces, make sure to quote the entire path.
Caveats
Please read the following usability issues before using NCM. These issues are listed in alphabetical order.
Administrative Settings - User Authentication Page Crypto Key Exception
It is possible that after upgrading to NCM 1.2, you will not be able to access any of the menu items under Administrative Settings. This is due to a corrupted encryption option in the site_options.rcx file.
Workaround:
Step 1
Go to the $NCM_HOME/jre directory.
Step 2
Backup the current site_options.rcx file.
Step 3
Open the site_options.rcx file and locate all encrypted text options by searching for EncryptedText.
Step 4
Remove the value for all encrypted text options if it is not empty. In the following example, you would delete the information between </comment> and </option>.
Before:
<option name="twist/password"><title>Twist Password</title><section>Cisco Server Automation System Authentication</section><size>30</size> <type>EncryptedText</type><comment>Web Services Data Access Engine Password for finding connected servers.</comment>encrypted:sQAHLgjGjdGIbvNB18NEoQ==</option>
After:
<option name="twist/password"><title>Twist Password</title><section>Cisco Server Automation System Authentication</section><size>30</size> <type>EncryptedText</type><comment Web Services Data Access Engine Password for finding connected servers.</comment></option>
Step 5
Save the file.
Step 6
Login to NCM.
Step 7
On the menu bar under Admin, select Administrative Settings and click User Authentication.
Step 8
Scroll down to the TACACS+ / RADIUS Authentication section.
Step 9
For the TACACS+ or RADIUS Secret option, enter the shared secret for the NCM host configured on the TACACS+ or RADIUS server.
Step 10
Scroll down to the Cisco Server Automation System Authentication section.
Step 11
For the Twist Password option, enter the SAS password to use when locating connected servers.
Step 12
Click Save.
Step 13
Click the Device Access tab.
Step 14
Scroll down to the Bastion Host Settings section.
Step 15
For the Default Bastion Host Password option, enter the password of the Bastion Host to use for Telnet and/or SSH access.
Step 16
Click Save.
Banner Handling Strings Require Device-specific Passwords
If you enter banner handling strings, Devices > Inventory > Edit > Show Device Access Settings (device-specific settings) > Setting > Banner skip regex option and enter common prompt strings, such as password or username, you cannot apply network-wide Password Rules to the device. If you do, the banner handling fails without generating any errors, and the device does not work with NCM device drivers. Tasks such as Snapshot and Driver Discovery do not work.
Workaround: Always use device-specific passwords on the Edit Device window.
Batch Insert ACL Line Option
When using the Batch Insert ACL Line option (Devices > New Device Task > Batch Insert ACL Line), the Task Options section on the New Task - Run Command Script window does not contain script content. While the Command Script to Run field correctly displays Cisco IOS Insert (or Remove) Line into (or from) ACL by handle, it does not present the script or script variables for execution until a device or device group for which the script supports is selected.
BayRS Device Can Lose Ability to Provide Snapshot
Occasionally, the BayRS device can enter a state in which it cannot provide a snapshot. Snapshot tasks fail with the following error message.
File retrieval error
Workaround: Rebooting the BayRS device restores the normal state on the device.
BayStack 450 Could Stop Responding to Telnet, SNMP, or ICMP
If you connect using a console to a BayStack 450, the allowed Telnet/SNMP Manager List is unexpectedly cleared out, indicating all management traffic is denied. This occurs when the device configuration file is downloaded repeatedly using TFTP. Nortel confirms this is an OS bug in some versions. The Nortel bug reference is CR 031215-85145.
Workaround: Do not snapshot more frequently than four times per day (the default). Be sure to turn off IGMP snooping if not in use. In case the BayStack 450 is unresponsive to Telnet, the switching function of the BayStack 450 is not affected. You should schedule a non-peak hour to reboot the device (or use terminal access to gain access).
Canceling or Deleting Tasks
Some NCM tasks will spawn external processes to run PERL or Expect scripts, or to run user-provided executables or shell scripts. Under certain circumstances, NCM may not be able to kill these external processes when the spawning task is cancelled or deleted. This could include scripts that spawn sub-processes or processes that are coded to catch kill signals.
Workaround: Manually stop the external process on the NCM server.
Cisco Banner Messages Special Characters
Cisco uses a superscript L (L) special character to begin and end banner messages in its configuration files. This character is not typically supported by XML. Consequently, when you create a policy enforcement rule, incorporating the L special character, you are able to export the policy, but not import the policy using this rule.
Workaround: You can manually edit the XML before importing the policy by adding a delimiting character before and after the banner, as long as the delimiting character does not occur in the banner itself.
Cisco Catalyst Switches
Catalyst switches running CatOS 8.3(3) could crash when you connect to them using SSHv2 (for example from an SSH client, such as SecureCRT or Putty). By default, NCM uses SSHv2 as the primary access method to network devices. Therefore, there is a substantial risk that a Catalyst switch running 8.3(3) could be reset when managed by NCM.
Workaround: Upgrade your Cisco Catalyst to CatOS 8.3(4). If this is not possible, edit your Catalyst devices running 8.3(3) in NCM to use only SSHv1 or Telnet for device access.
Command Line Interface: connect Command
The connect command in the NCM Proxy now accepts a device ID. This is needed because device IP Addresses are no longer required to be unique. If you pass an invalid device ID, such as an ID that is not a number, with the connect command the NCM Proxy session is abruptly terminated.
Workaround: Reconnect to the NCM Proxy and enter a valid device ID.
Command Line Interface: Set Telnet or SSH Client Width to 500
The NCM CLI has very wide output. For maximum ease in viewing the data, set your client's buffer width to 500.
Console Server: SSH Access is not Supported
NCM does not support console server access using SSH. If you use a console server to access a device, you must use the Telnet connectivity. In other words, in the New Device window /Edit Device window, if Use to access device is checked in the Console Server Information section, you should make sure that the Telnet option in the Connection Information section is also checked.
Deploy to Startup Config and Reboot not Supported Using SNMP
NCM can deploy a configuration file to the startup configuration and reboot the device using the command line only. If the device is configured for SNMP access only (see the Device Driver Reference for Network Compliance Manager for Network Compliance Manager), deploy startup and reboot will fail.
Detect Network Devices Task
The NCM system prevents you from inadvertently running more than one Detect Network Devices task concurrently. Although the Detect Network Devices task generates only a minimal level of traffic, NCM provides this protection to help minimize additional traffic when running duplicate or additional Detect Network Devices tasks simultaneously. If a second or third Detect Network Devices task is scheduled while an earlier Detect Network Devices task is running, NCM will place the new task(s) in the Waiting state. The task(s) will run individually after the first Detect Network Devices task has completed.
Diagnostics: When to Run ICMP Tests
Use ICMP tests only to verify connectivity occasionally or after a change. They are not a replacement for monitoring software. You should schedule ICMP tests no more than once per 10 minutes.
Diagramming
NCM applies an absolute value for the text height attribute for interface and port labels shown in Visio diagrams. When the Visio VDX file is loaded, Visio assigns an incorrect formula to the text height attribute. As a result, when you have more than two lines of annotated text, such as a label, for an interface or port and you attempt to copy and paste, the label of the new interface or port is displayed improperly and could hide the interface or port icon.
Workaround: Click the Text Tool option on the Visio tool bar and move the label so as to expose the interface or port icon.
Displaying Diagnostics
Most NCM diagnostics are stored in text format. For a list of NCM diagnostics, from the Reports drop-down menu, select Search For and click Diagnostics. The following NCM diagnostics, however, are stored in binary format, and therefore are not searchable:
•
NCM Module Status
•
NCM Routing Table
•
NCM OSPF Neighbors
•
NCM Interfaces
•
NCM Flash Storage Space
Workaround: Because the issue is that built-in diagnostics are not stored as clear text, you can create a custom diagnostic that performs the appropriate command (for instance, Show Interfaces for the equivalent of Module Status). As a result, the custom diagnostic will be searchable.
Distributed System Performance
When running a Distributed System, if you are deleting many objects simultaneously, the system may take a while to push transactions for large delete operations.
Duplicate IP Addresses with Multiple Sites
If your system is configured with multiple Sites in different Realms, you could see duplicate IP addresses if you select the Multiple Devices/Groups option on a New Task window when browsing the Inventory Group using the Device Selector.
Workaround: Using the Device Selector, browse to devices using the specific Site Group.
Extreme Devices: Configuration Comments Can Cause Misconfiguration
On Extreme devices, adding inline comments between multi-line commands, such as user account commands or set banner commands, can cause serious problems if the resulting configuration is deployed.
Workaround: Do not add inline comments between multi-line commands. Add comments on the line above the start of a command.
Installing NCM on Linux
When installing NCM on a Linux platform, the install might fail because there is no access to the MySQL database.
Workaround: When installing NCM on a Linux platform, perform the following steps prior to starting the installation.
Step 1
1. Open the /etc/hosts file.
Step 2
2. Change 127.0.0.1 localhost.localdomain localhost to: 127.0.0.1 localhost.
3. Save and close the /etc/hosts file.
If you have already started the installation, use the Linux command line to run the following commands:
#mysql -h <device ip - not 127.0.0.1> -u root mysql
mysql> GRANT ALL PRIVILEGES ON *.* TO root@localhost.localdomain IDENTIFIED BY '<password&

